New Bipolar Boost TX and Front End

[KingJL]

I have modified the rear of the main enclosure to allow for better access and assembly of the power/battery module.
Prior to this change, the power module needed to be installed before the 2 halves of the enclosure were screwed together, not a good solution, if repairs were needed.

[ATTACH]47226[/ATTACH]

Also JL, there is room for another PCB 25 x 50mm in this area [ATTACH]47227[/ATTACH] and I am fairly certain there is room on the power module to mount same components as an option.

While geotech1 was inaccessible, I sent you an email with kicad files and datasheet for he probe relay. The relay pads are probably about 0.25mm too large. I have already reduced them to 0.75 mm on the kicad footprint in the files that I sent you. I think the ideal size would be 0.70mm with the annular ring diameter of 1.4mm. The actual pin measures 0.020" (about 0.5mm). They also make a version of the relay for surface mount, but there is really nothing to be gained as the physical size of the relay is the same. In fact it may take more board space as the leads for the surface mount extend 1 mm (each side) outside the case profile.

[Mdtoday]

I have modified the rear of the main enclosure to allow for better access and assembly of the power/battery module.
Prior to this change, the power module needed to be installed before the 2 halves of the enclosure were screwed together, not a good solution, if repairs were needed.



Also JL, there is room for another PCB 25 x 50mm in this area and I am fairly certain there is room on the power module to mount same components as an option.

[Mdtoday]

I received the relays today. I will provide the dimensions and maybe a prelim schematic for the probe relay circuit. I do not know where we are going to put it. It will involve the relay, 9 connections (3.3v, GND, control line from PMOD, and 2 connections each for probe, search coil, TX board). It also has minimal circuitry consisting of damp resistor for the unused coil, current limit resistor for Pmod input, and 2n3904 relay control transistor.

Excellent, perhaps, we can include that on the power board as an option and still do a separate small board, this would probably depend on coil wire separation from switch-mode supplies.
If on its own Pcb, it could be mounted on rear main electronics wall.

[Mdtoday]

I have used the MT3608 modules in my latest Magnetic Viscosity Meter and encountered a couple of problems. Yes, it is boost only and can only be set to higher voltages than the input. My first problem was noise, both radiated and conducted. The conducted noise was reduced to an acceptable level by adding 10uF ceramic through hole capacitors across input and output pads. There are supposedly 22uF already on the board, but I found on a couple of boards they were only 10uF so it is worth checking. Radiated noise comes from the inductor on the module and I had to put the whole module in a small metal screening box to prevent this. Noise levels appear to depend on the amount of boost required and I ended up with a 11.1V three cell Li-ion battery back by Ansmann and boost up to 15V which the TX on the MVM requires. Other lower voltages are obtained from linear regulators. I blew a couple of MT3608's by having it charge a TX reservoir capacitor on switch on, so be careful in this area.

Battery protection is important as I found out the hard way. Inadvertently leaving the MVM on caused unrecoverable death to a couple of earlier single cell versions. As the input voltage falls the module attempts the keep the output voltage constant, hence more and more current is drawn from the battery, hastening its demise. The built in protection on the Ansmann pack just switches the battery out of circuit when the voltage falls to around 2.9V per cell as well as protecting from overcharge current.

Yes, thanks for the input on JL's project Eric and good points, perhaps we can change the inductor footprint on the on board circuits to a shielded version which is what I generally use and add some touch points for a tin shield on top side ground plane just in case.
nickle spray on the internal inside wall between battery bay and main electronics enclosure would help too.
I will be able to test the whole power module including battery protection sometime this week.

Cheers

Mdtoday

[KingJL]

Here is the latest schematic version of the Power/Bat protection PCB (untested..still waiting on parts and modules)

[ATTACH]47221[/ATTACH]

I will post the final Kicad files once all is tested and confirmed working ok.

I received the relays today. I will provide the dimensions and maybe a prelim schematic for the probe relay circuit. I do not know where we are going to put it. It will involve the relay, 9 connections (3.3v, GND, control line from PMOD, and 2 connections each for probe, search coil, TX board). It also has minimal circuitry consisting of damp resistor for the unused coil, current limit resistor for Pmod input, and 2n3904 relay control transistor.

[KingJL]

I have used the MT3608 modules in my latest Magnetic Viscosity Meter and encountered a couple of problems. Yes, it is boost only and can only be set to higher voltages than the input. My first problem was noise, both radiated and conducted. The conducted noise was reduced to an acceptable level by adding 10uF ceramic through hole capacitors across input and output pads. There are supposedly 22uF already on the board, but I found on a couple of boards they were only 10uF so it is worth checking. Radiated noise comes from the inductor on the module and I had to put the whole module in a small metal screening box to prevent this. Noise levels appear to depend on the amount of boost required and I ended up with a 11.1V three cell Li-ion battery back by Ansmann and boost up to 15V which the TX on the MVM requires. Other lower voltages are obtained from linear regulators. I blew a couple of MT3608's by having it charge a TX reservoir capacitor on switch on, so be careful in this area...

Eric, I remembered your experience with the '3608. I included a lm317 current limiter circuit in the source path of the 5V (where there is a 4700uF resevoir). Thank you for your comments. BTW, I have been ordering enough parts to populate 2 sets of boards. I owe you so much for the opportunity to study and learn from the Vallon.

[Ferric Toes]

Here is the latest schematic version of the Power/Bat protection PCB (untested..still waiting on parts and modules)

[ATTACH]47221[/ATTACH]

I will post the final Kicad files once all is tested and confirmed working ok.

I have used the MT3608 modules in my latest Magnetic Viscosity Meter and encountered a couple of problems. Yes, it is boost only and can only be set to higher voltages than the input. My first problem was noise, both radiated and conducted. The conducted noise was reduced to an acceptable level by adding 10uF ceramic through hole capacitors across input and output pads. There are supposedly 22uF already on the board, but I found on a couple of boards they were only 10uF so it is worth checking. Radiated noise comes from the inductor on the module and I had to put the whole module in a small metal screening box to prevent this. Noise levels appear to depend on the amount of boost required and I ended up with a 11.1V three cell Li-ion battery back by Ansmann and boost up to 15V which the TX on the MVM requires. Other lower voltages are obtained from linear regulators. I blew a couple of MT3608's by having it charge a TX reservoir capacitor on switch on, so be careful in this area.

Battery protection is important as I found out the hard way. Inadvertently leaving the MVM on caused unrecoverable death to a couple of earlier single cell versions. As the input voltage falls the module attempts the keep the output voltage constant, hence more and more current is drawn from the battery, hastening its demise. The built in protection on the Ansmann pack just switches the battery out of circuit when the voltage falls to around 2.9V per cell as well as protecting from overcharge current.

Great work on this project between the two of you.

Eric.

[Mdtoday]

Here is the latest schematic version of the Power/Bat protection PCB (untested..still waiting on parts and modules)

KingJL_PWR_RevB.pdf

I will post the final Kicad files once all is tested and confirmed working ok.

[Mdtoday]

I have added an option to the power board to mount either MT3608 modules or mounting the same circuit idea directly on the board.



There were 2 reasons for this, one was to allow for easy preset of the output voltages with vertical mount trimpots, the other was the delivery time 3>6weeks for the modules and the cost here in Australia from local sellers has been rising.
With plenty of space on the PCB, I thought it was worth it and the actual low cost of parts from LCSC made it more so.

[Mdtoday]

Link now shows!!!

Not sure what went wrong there!

Here is the general layout for added protection circuit hopefully shows up.

[KingJL]

??????

Link now shows!!!

[KingJL]

Here is a draft of cell protection circuit,

??????

[Mdtoday]

On the Power PCB my thoughts are to include UV/reverse polarity protection circuits for each 18650 cell to prevent any nasty problems occurring.
.

Here is a draft of cell protection circuit, It's a work in progress

IC1 and IC2 along with Q1,Q2 respectively provide the protection for each 18650 cell.
D1 and D2 provide the reverse polarity protection in case either battery is inserted the wrong way.
This will cause the UV/OV points to move but I think this can be corrected somewhat by inserting a voltage divider in place of R3 and R4
D3 and D4 provide the isolation for each protection circuit.
D5 + D1/D3 or D2/D4 in series should drop the voltage enough to the MT3608 module U3 to work at a lower voltage required to boost up to 3v3... it's not tested yet so we will see once I get delivery of some modules...

D3, D4,D5 may have to be standard diodes to get the required voltage drop to U3

The MT3608 modules work down to 2v

Additional cost of the protection circuit is around $4-$5

[Mdtoday]

On the Power PCB my thoughts are to include UV/reverse polarity protection circuits for each 18650 cell to prevent any nasty problems occurring.
Probably using something like the TI BQ279700, Seiko S-8200 or Fortune DW01-P devices.
It adds cost but with Li power cells, safety is always a concern.

[Mdtoday]

so far in all of my projects, I have only used them with 3 and 4.5 volt inputs... and I had not yet used them to generate 3.3V.,, it was always 5V and higher. I probably need to hook one up and test for generating 3.3V.

Thanks, that makes sense, I should receive the order of MT3608 modules in a day or two and can check, if you are busy with code.